First responses delivered on time, up from 23% before the transformation.
Cases fully resolved within one day, up from 15% before introducing the new service model.
Of chat conversations resolved without a human agent, powered by HALO
BERG is a Dutch outdoor toy maker known for trampolines, pedal go-karts and play equipment built to last, some of its products are still in use after 30 to 40 years. For most of its history BERG sold business-to-business, but in 2019 it began selling directly to consumers. In just five years it grew from three webshops to seventeen, and today it serves consumers across 17 countries and multiple languages, with offices in the Netherlands, Switzerland and the United States.
At heart, it remains a personal brand. "Personal" is one of BERG's core values: the company wants every customer, from a long-standing retailer to a parent buying their child's first trampoline, to feel genuinely looked after, like part of the family. With an in-house customer service team and a strongly seasonal business, that close, human relationship with customers is central to how BERG sees itself.
BERG had made the leap from B2B to selling directly to consumers, but its systems hadn't kept up. A simple ticketing system left agents with no overview, no history and no shared data, so as consumer volume grew across seventeen markets, customer contact began to break down. In peak season, a first response could take up to two weeks, at one point BERG even switched off its phone lines to clear the backlog.
The fix wasn't a single tool but a disciplined sequence: BERG first brought its scattered channels onto one connected platform, then made performance measurable, and only then began to automate. The result is a service operation that scales with the business instead of breaking under it.
BERG's move from B2B to B2C changed who it served almost overnight and its systems and processes weren't built for it. A setup made for a handful of business relationships couldn't absorb a flood of individual consumers across seventeen markets.
At peak, the backlog ran into the hundreds of unresolved cases a month. Customers who got no answer simply tried again elsewhere, email, phone, social, chat, so one question landed in several places at once, multiplying the work and pulling the whole organization in. First responses could take up to two weeks.
Underneath sat a simple ticketing system and disconnected channels: agents switched tools constantly, with no customer history, no shared data and no ownership. The real sting was the gap between the product and the experience around it and BERG knew it wasn't sustainable.
We were just fighting to get things done. We had to act to become future-proof. Thecla Huf, Manager Customer Service, BERG
BERG started where the pressure was highest, and built up in clear steps. First it connected Mobile Service Cloud, bringing every channel, conversation and customer history into one shared inbox. For the first time agents had a single view of their workload and the team had the dashboards and data to see exactly where the pressure came from, and to set clear targets for response and resolution times.
With that foundation in place, BERG turned to automation. Using HALO, CM.com's agentic AI platform, it built AI agents that don't just answer but act handling routine questions end to end across chat and email, and freeing the team for the cases that genuinely need a human.
Platform: Mobile Service Cloud and Mobile Marketing Cloud, every channel, customer history and dashboard in one place, connected to Magento and Dynamics 365
Channels: Webchat, email, voice and social media handled from a single agent workspace
Automation: HALO AI agents resolving routine questions end to end across chat and email
Business benefit: The overview, ownership and data to hold response times steady as volumes grow
With response times under control, the team could finally ask where its volume came from. The answer was clear: about one in five questions in the Netherlands and Germany was about spare parts, no surprise for products that last decades. Answering them meant slow, repetitive digging through systems by article number, so BERG built a dedicated HALO agent to handle the whole flow, in both webchat and email.
Result: Monthly spare-part questions in the Netherlands and Germany dropped from around 1,070 to 150, an 86% reduction.
< See the Spare-Agent in action on the BERG website
Guides customers to find the article or batch number
Searches the ERP system for matching spare parts
Presents the matching parts in a selectable carousel
Generates a direct URL to the chosen product
Not every question is a problem, many are about choosing. Customers often asked which trampoline or go-kart suited their child's age, garden or use, so BERG taught the bot to give that advice right in the chat: true to its "personal" value, with a human only ever one click away.
Result: 84.5% of chat conversations now resolve without a human agent, including guided ones like this.
See how HALO is helping with choosing advice on the BERG website>
Asks about age, available space and intended use
Narrows the range to the best-fit models
Explains the differences between product lines
Hands over to a human the moment it's wanted
Getting the volume under control also meant changing where customers landed in the first place. BERG made chat the front door of its service operation: a full-width inline chat widget in the Netherlands and Germany, with the other channels moved further down the page and a human still only one click away, true to the "personal" core value. The shift worked. Chat sessions grew sixfold, from 332 in March 2025 to 2,044 a year later and yet only 15.5% needed a human agent. HALO resolved the rest.
That changed what the team actually does all day. With automation absorbing the routine volume, agents and specialists finally had room to look up instead of just keeping up. They could dig into the tagged data to see where questions really came from, improve the knowledge and content behind the answers, clearer product pages, better information, fewer reasons to get in touch at all and decide what to automate next. The job began shifting from answering the same questions over and over to improving the system that answers them, and from reacting to volume to removing its causes.
The results speak for themselves.
Of all chat conversations now resolved automatically
Fewer messages needed to close a case (2.72 → 1.65)
Fewer spare-part questions reaching the team, now handled by one agent
Fewer invoice requests after a successful self-service fix
These gains compound into something the team feels every day. Fewer messages to close a case means far less back-and-forth between agents and a lighter load when the season peaks and as avoidable queries fall away, from lead-time questions (−55%) to invoice requests (−70%), the same team simply handles more.
For a company chasing ambitious international growth, that headroom is the prize. BERG can scale into new markets and ride out seasonal peaks on the strength of its platform and data rather than its headcount and customers finally get service that matches the quality of the products.
With the foundation in place and automation proving itself, BERG's focus is shifting to the next phase of scale:
Rolling out the Spare Part Agent in other countries
Reaching a 95% on-time first-response rate through automation
Adding CSAT and NPS to the KPI set
Launching more HALO agents, guided by data on what to automate next
There is no growth without innovation, so you make sure everyone is aligned and informed, and you keep moving forward. Thecla Huf, Manager Customer Service, BERG
Thecla: "You have to dare to do it. Sometimes you can elaborate for the tenth time, but at some point it's time to take action and that's why we're here with these results."
Select a region to show relevant information. This may change the language.
